Abstract An experiment will be described which shows that DNA polymerase helps select the base in DNA replication. After elucidating the structure of DNA (1953a), Watson and Crick proposed a template mechanism for the replication of DNA (1953b). This hypothesis invokes the same hydrogen bonds which hold the finished DNA strands together as the agents for the selection of precursor nucleotides to complement those of the parental DNA strand. The enzyme for this polymerization process was subsequently discovered, Kornberg (1960) , and was not previously known to play a selective role. Nor was it implicated in mutagenesis. In another selective polymerization directed by a nucleic acid, protein synthesis, the ribosomes-which may be considered as enzymes-affect the specificity of the process, Gorini and Kataja (1964) . This suggested that the enzymes involved in nucleic acid synthesis might also affect the specificity. Since enzymes may change their substrate specificity when their gene is mutated, Hotchkiss and Evans (1960) , a mutant DNA polymerase might therefore cause errors in DNA synthesis. These errors can be detected as an increase in the frequency of mutations. When a mutant DNA polymerase was discovered a direct test became possible. This discovery was made by Epstein and Edgar and their colleagues (1963) who mapped the genes of coliphage T4. They have found about 70 genes, of which several are involved in DNA replication. One of these, gene 43, (Edgar et al , 1964) has been identified as a structural gene of DNA polymerase by de Waard et al (1965) . Dr. Edgar generously gave us thirteen temperature sensitive mutants of this gene. We have studied the effect of two ts alleles of this gene on mutagenesis.